At a Glance · AI-generated

Meta launched its Muse AI assistant on September 22 but immediately disclosed a critical zero-day vulnerability that allowed attackers to control the agent. The exploit leveraged Muse's elevated system privileges to compromise its operations. Meta released a patch the same day to address the security flaw. The incident highlighted risks associated with deploying highly privileged AI systems before thorough security testing.